2021 International Conference on Medical Imaging Physics and Engineering and the 10th National Annual Meeting of Medical Imaging Physics (ICMIPE 2021)

46 / 2021-10-14 10:28:02

Image domain material decomposition algorithm by plug-and-play framework for dual-energy CT sparse-view scanning

Dual-energy computed tomography,plug-and-play,FFDnet denoiser,material decomposition

Dual-energy computed tomography (DECT) has high application prospects in distinguishing and quantifying materials. However, DECT requires at least two full-angle scans at different energies. In this paper, in order to reduce the radiation dose, we use the plug-and-play (PnP) framework to obtain high-quality decomposed materials under sparse angle scanning. Specifically, we design the PnP framework as a combination of the FFDnet denoiser prior and the least square estimation model to suppress artifacts. To verify the applicability of the proposed method in the clinical environment, we simulated the chest cavity for experiments. The results show that the proposed methods can achieve high-quality basis material decomposition under different sparse views.